Dravidian Languages Research Articles

Double-finite Serial Verbs in Sanskrit and later Indo-Aryan

Abstract Steever (1988. The serial verb formation in the Dravidian languages. Delhi: Motilal Banarsidass) introduced to South Asian linguistics the notion “Serial Verb” – a Dravidian construction in which two morphologically finite verbs agree in person and number features in the same sentence. I show that similar “double-finite” structures are found in Sanskrit/Indo-Aryan, although details differ. One double-finite construction, inherited from Proto-Indo-European, does not seem to have continued in the post-Vedic language, but others do, and similar structures occur in Modern Indo-Aryan languages. While the historical developments are relatively clear for early Indo-Aryan, from Vedic down to Pāli, there are questions about whether the double-finite constructions of Modern Indo-Aryan are inherited , since comparable structures have not been noted in late Middle Indo-Aryan Apabhraṁśa, the near-ancestor of (peninsular) Modern Indo-Aryan. In the conclusion I discuss alternatives to Modern Indo-Aryan inheritance, including areal and typological factors and the proposal by Milizia (2018. Sul presente neoindoarico a doppio verbo finito. Atti del Sodalizio Glottologico Milanese 12(n.s.). 93–111) that the perhaps most widespread Modern Indo-Aryan double-finite construction results from analogical developments. (Unfortunately, the latter proposal does not account for the other Modern Indo-Aryan double-finite constructions.)

Enhancing Neural Machine Translation Quality for Kannada–Tulu Language Pairs through Transformer Architecture: A Linguistic Feature Integration

The rise of intelligent systems demands good machine translation models that are less data hungry and more efficient, especially for low- and extremely-low-resource languages with few or no data available. By integrating a linguistic feature to enhance the quality of translation, we have developed a generic Neural Machine Translation (NMT) model for Kannada–Tulu language pairs. The NMT model uses Transformer architecture and a state-of-the-art model for translating text from Kannada to Tulu and learns based on the parallel data. Kannada and Tulu are both low-resource Dravidian languages, with Tulu recognised as an extremely-low-resource language. Dravidian languages are morphologically rich and are highly agglutinative in nature and there exist only a few NMT models for Kannada–Tulu language pairs. They exhibit poor translation scores as they fail to capture the linguistic features of the language. The proposed generic approach can benefit other low-resource Indic languages that have smaller parallel corpora for NMT tasks. Evaluation metrics like Bilingual Evaluation Understudy (BLEU), character-level F-score (chrF) and Word Error Rate (WER) are considered to obtain the improved translation scores for the linguistic-feature-embedded NMT model. These results hold promise for further experimentation with other low- and extremely-low-resource language pairs.

The Qur’an in Malayalam: Contextualising Early iṣlāḥī Translations

This article is an exploration of the texts and historical contexts of the early translations of the Qur'an into the South Indian language of Malayalam. Attempts at Malayalam translations of the Qur'an started in the 1930s. The present study deals with the translation projects that emerged between the 1930s and 1965. It showcases the situatedness of these translations in the colonial experiences of the region, as well as in the discursiveness of global Islamic trends. It also looks at the theological controversies these translations evoked within the Muslim community.

Linguistic discussions of Urdu Language

The beginning and evolution of Urdu language is a controversial issue in our country. The research and editing of linguists regarding the early characteristics of Urdu language and its development is still going on. First, European historians researched and edited linguistics and presented their ideas. These theories were an early reference, hence their importance to Islam. Subsequent researchers and linguists carried on the same work. The beginning of the twentieth century marks the beginning of a vernacular period of linguistic research. During this period, there is a tendency to assign Urdu to a particular place or place. Many researchers have traced the traces of Urdu in the Mughal period. Among them, Mera Man, Sir Syed Ahmad Khan, Maulana Sohbai, Shamsullah Qadri and Maulana Muhammad Hussain Azad are worth mentioning. And some say it is a Dravidian language and some say Urdu is the daughter of Braj Bhasha. There are many theories about the origin of the Urdu language, but it is not entirely correct to connect it with any one region.

Sentiment analysis of code-mixed Dravidian languages leveraging pretrained model and word-level language tag

Sentiment analysis of code-mixed Dravidian languages leveraging pretrained model and word-level language tag

Harnessing heterogeneity: A multi-embedding ensemble approach for detecting fake news in Dravidian languages

Fake news detection in low-resourced languages has become a critical challenge in our information-driven society. Contextualized word embeddings can improve the accuracy of fake news detection in low-resource languages compared to traditional word embedding methods. However, seizing the complex relationships among words in news articles is challenging when using a specific contextualized word embedding strategy. To address this issue, we propose a computationally efficient heterogeneous ensemble-based Dravidian fake news detection framework that utilizes gradient accumulation to enhance base-level and meta-level classifiers. This framework employs three optimized BiLSTM base-level classifiers, each paired with different contextualized word embeddings: multilingual BERT, XLM-RoBERTa, and MuRIL. The predictions from these base classifiers are integrated using an attention mechanism and fed into an optimized LSTM meta-classifier for final classification, ensuring a comprehensive and context-aware decision-making process. On the Dravidian_Fake dataset (Telugu, Kannada, Tamil, and Malayalam), our framework achieves significantly higher accuracy and F1 scores compared to baselines, demonstrating its effectiveness in capturing complex word relations and robustly detecting fake news in low-resource settings.

A novel socio-pragmatic framework for sentiment analysis in Dravidian–English code-switched texts

The advent of global digital communication has accentuated the complexity of multilingual code-switching, presenting formidable challenges for sentiment analysis by necessitating nuanced interpretation of linguistic and socio-pragmatic cues. This research introduces MultiSwitchNet, a pioneering sentiment analysis model tailored for the intricate dynamics of Dravidian languages mixed with English code-switched texts. MultiSwitchNet leverages the linguistic prowess of XLM-RoBERTa, combined with socio-pragmatic insights through a novel Socio-Pragmatic Embedding Layer (SPEL) utilizing DistilBERT. Evaluated on the Dravidian Code-Mixed Dataset, MultiSwitchNet achieved superior performance metrics, including an accuracy of 72.3%, precision of 72.9%, recall of 72.1%, and an F1-score of 72.5%. Its effectiveness extends across varying degrees of code-switching complexity and demonstrates remarkable consistency in sentiment prediction across semantically similar yet syntactically diverse sentences. Statistical analyses, including ANOVA and post hoc tests, underscore the model’s significant outperformance compared to both baseline and state-of-the-art models, highlighting the indispensability of integrating socio-pragmatic awareness for accurate sentiment analysis. MultiSwitchNet’s innovative approach sets a new benchmark for sentiment analysis in multilingual contexts and opens avenues for future research into models capable of discerning the rich socio-pragmatic fabric of global digital communication.

Does D Select the CP in Light Verb Constructions? A Reply to Hankamer and Mikkelsen 2021

In this response to Hankamer and Mikkelsen (H&M) 2021, we clarify the meaning of the definite article the in out-of-the-blue occurrences of English light verb constructions such as make the claim. While H&M view these as purely uniqueness-requiring determiners (as opposed to purely anaphoric ones in non−light verb constructions), we classify them as instances of Carlsonian weak definites (e.g., Carlson et al. 2006), which presuppose neither uniqueness nor anaphoricity. This clarification casts doubt on H&M’s idea that the D in light verb constructions selects the descriptive CP, which serves to uniquely specify the intended referent. Crosslinguistic data (from the Dravidian language Kannada) also serve to question the relationship between D and CP claimed by H&M. We conclude that D does not directly select the CP in light verb constructions. Nevertheless, these CPs do empirically behave like syntactically selected complements. We suggest that this is because, if the light verb construction is taken as a whole—that is, the light verb + the weak definite nominal—its complement selection properties altogether parallel those of the corresponding lexical verbs. We propose a compositional analysis that reinforces this conclusion.

Investigating linguistic and genetic shifts in East Indian tribal groups

South Asia is home to almost a quarter of the world's total population and is home to significant ethnolinguistic diversity. Previous studies of linguistic and genetic affiliations of Indian populations suggest that the formation of these distinct groups was a protracted and complex phenomenon involving multiple waves of migration, cultural assimilation, and genetic admixture. The evolutionary processes of migration, mixing and merging of populations thus impact the culture and linguistic diversity of different groups, some of which may retain their linguistic affinities despite genetic admixture with other groups, or vice versa. Our study examines the relationship of genetic and linguistic affinities between Austroasiatic and Indo-European speakers in adjacent geographical regions of Eastern India. We analyzed 224 mitogenomes and 0.65 million SNP genotypes from 40 unrelated individuals belonging to the Bathudi, Bhumij, Ho, and Mahali ethnic groups from the Eastern Indian state of Odisha. These four groups are speakers of Austroasiatic languages who have adopted elements from Indo-European languages spoken in neighbouring regions. Our results suggest that these groups have the greatest maternal genetic affinity with other Austroasiatic-speaking groups in India. Allele frequency-based analyses, genome-wide SNPs, haplotype-based methods and IBD sharing further support the genetic similarity of these East Indian groups to Austroasiatic speakers of South Asia rather than regional populations speaking Indo-European and Dravidian languages. Our study shows that these populations experienced linguistic mixing, likely due to industrialization and modernization that brought them into close cultural contact with neighbouring Indo-European-speaking groups. However, linguistic change in these groups is not reflected in genetic mixing in these populations, as they appear to maintain strict genetic boundaries while simultaneously experiencing cultural mixing.

Telugu language hate speech detection using deep learning transformer models: Corpus generation and evaluation

In today's digital era, social media has become a new tool for communication and sharing information, with the availability of high-speed internet it tends to reach the masses much faster. Lack of regulations and ethics have made advancement in the proliferation of abusive language and hate speech has become a growing concern on social media platforms in the form of posts, replies, and comments towards individuals, groups, religions, and communities. However, the process of classification of hate speech manually on online platforms is cumbersome and impractical due to the excessive amount of data being generated. Therefore, it is crucial to automatically filter online content to identify and eliminate hate speech from social media. Widely spoken resource-rich languages like English have driven the research and achieved the desired result due to the accessibility of large corpora, annotated datasets, and tools. Resource-constrained languages are not able to achieve the benefits of advancement due to a lack of data corpus and annotated datasets. India has diverse languages that change with demographics and languages that have limited data availability and semantic differences. Telugu is one of the low-resource Dravidian languages spoken in the southern part of India.In this paper, we present a monolingual Telugu corpus consisting of tweets posted on Twitter annotated with hate and non-hate labels and experiments to provide a comparison of state-of-the-art fine-tuned deep learning models (mBERT, DistilBERT, IndicBERT, NLLB, Muril, RNN+LSTM, XLM-RoBERTa, and Indic-Bart). Through transfer learning and hyperparameter tuning, the models are compared for their effectiveness in classifying hate speech in Telugu text. The fine-tuned mBERT model outperformed all other fine-tuned models achieving an accuracy of 98.2. The authors also propose a deployment model for social media accounts.

AI Based Multilingual College Enquiry Voice Bot Using Python

Abstract: This paper addresses the growing need for seamless cross-language communication in our global society. Leveraging AI, it offers a versatile conversational agent capable of understanding and responding to inquiries in multiple languages. Through speech recognition and advanced NLP, the system interprets user intent and retrieves relevant information from a college database. Integrated language translation services enable communication in various languages namely-Kannada, English, Tamil and Telugu enhancing accessibility. Responses are dynamically translated into the user's preferred language using text-to-speech conversion. This project demonstrates AI's potential in bridging linguistic barriers and improving access to educational resources. Future enhancements could include personalized responses and expanded language support in five South Indian languages, further enriching the user experience and solidifying the bot's role as a valuable educational tool.

Handwritten Telugu Character Recognition & Signature Verification

Behaviour reputation stands as one of the earliest applications in sample reputation. While spotting handwritten characters is an clean venture for humans, it is a formidable task for computer structures. Optical Character Recognition (OCR) is an crucial answer primarily based on optical systems, which enables automatic reputation of scanned and digitized characters This paper explores into optical man or woman popularity strategies in particular developed for handwriting Telugu within the characters. Telugu, a Dravidian language spoken especially in Andhra Pradesh and Telangana, India, offers precise challenges because of its complex alphabet Basic parts of Telugu script together with "vattu" which stands for vowels and "gunitalu" which means that tone the complicated syllables add to the complexity. Combining OCR strategies with Harris corner popularity, the paper affords insights into the accuracy and efficiency of handwritten Telugu person reputation and the fidelity of handwriting This have a look at contributes to the development of character reputation in particular on in complex written languages ​​which include Telugu and gives realistic explanations for handwriting verification processing. Key Words: Optical Character Recognition (OCR), Telugu, Vattu, Gunitalu, Harris corner detection, Handwritten Character Recognition, Signature Verification.

An Empirical Analysis of Language Detection in Dravidian Languages

Objectives: Language detection is the process of identifying a language associated with a text. The proposed system aims to detect the Dravidian language that is associated with the given text using different machine learning and deep learning algorithms. The paper presents an empirical analysis of the results obtained using the different models. It also aims to evaluate the performance of a language agnostic model for the purpose of language detection. Method: An empirical analysis of Dravidian language identification in social media text using machine learning and deep learning approaches with k-fold cross validation has been implemented. The identification of Dravidian languages, including Tamil, Malayalam, Tamil Code Mix, and Malayalam Code Mix, is performed using both machine learning (ML) and deep learning algorithms. The machine learning algorithms used for language detection are Naive Bayes (NB), Multinomial Logistic Regression (MLR), Support Vector Machine (SVM), and Random Forest (RF). The supervised Deep Learning (DL) models used include BERT, mBERT and language agnostic models. Findings: The language agnostic model outperform all other models considering the task of language detection in Dravidian languages. The results of both the ML and DL models are analyzed empirically with performance measures like accuracy, precision, recall, and f1-score. The accuracy associated with different machine learning algorithms varies from 85% to 89%. It is evident from the experimental result that the deep learning model outperformed with an accuracy of 98%. Novelty: The proposed system emphasizes on the use of the language agnostic model to implement the process of detecting Dravidian languages associated with the given text which provides a promising result of 98% accuracy which is higher than the existing methodologies. Keywords: Language, Machine learning, Deep learning, Transformer model, Encoder, Decoder

Fake news detection in Dravidian languages using multiscale residual CNN_BiLSTM hybrid model

Fake news detection is the process of identifying news that contain purposeful misinformation disseminated through traditional news sources or social media platforms. Detecting fake news in Dravidian languages poses unique challenges due to the linguistic diversity and limited resources available for these languages. In this study, we propose a novel approach for fake news detection in Dravidian languages, leveraging contextualized word embeddings from MuRIL. MuRIL provides contextual embeddings that capture nuanced linguistic nuances, making it well-suited for the complexities of Dravidian languages. The proposed hybrid model combines multiscale residual CNN and BiLSTM layers to capture local and global dependencies in textual data effectively. The multiscale architecture allows the model to extract features at various levels of granularity while the BiLSTM layer captures long-range dependencies and contextual information. We performed extensive experiments on diverse datasets in Dravidian languages to evaluate the proposed approach. The results demonstrate the effectiveness of our hybrid model in discerning between genuine and fake news articles.

“Deep Learning Based Classification of Single-Hand South Indian Sign Language Gestures”

Gesture recognition is a branch of computer science and language technology dedicated to utilizing mathematical algorithms for the analysis of human gestures. Within the realm of non-verbal communication, the pivotal role of human arm movements and gestures remains a focal point. This research introduces advanced multi-stream deep transfer learning models tailored for identifying signs from South Indian languages, specifically Kannada, Tamil, and Telugu. The primary aim is to offer support to individuals encountering speech disorders or disabilities. The key deep transfer learning models utilized include Inception-V3, VGG-16, and ResNet-50, which have been modified and improved to attain heightened classification efficacy. The dataset comprises 35,000 images capturing single-hand gestures. In the realm of models, Inception-V3 exhibits the utmost test accuracy, standing at 91.45%, alongside a validation accuracy of 93.45% when tasked with the classification of single-hand gesture images across thirty-five distinct categories. The importance of this study lies in its prospective utility for creating an automated system that can support and improve the functional capabilities of individuals facing speech disorders or disabilities.

Deep Learning Approaches for Age-based Gesture Classification in South Indian Sign Language

This study focuses on recognizing and categorizing South Indian Sign Language gestures based on different age groups through transfer learning models. Sign language serves as a natural and expressive communication method for individuals with hearing impairments. This study intends to develop deep transfer learning models, namely Inception-V3, VGG-16, and ResNet-50, to accurately identify and classify double-handed gestures in South Indian languages, like Kannada, Tamil, and Telugu. A dataset comprising 30,000 images of double-handed gestures, with 10,000 images for each considered age group (1-7, 8-25, and 25 and above), is utilized to enhance and modify the models for improved classification performance. Amongst the tested models, Inception-V3 achieves the best performance with a test precision of 95.20% and validation accuracy of 92.45%, demonstrating its effectiveness in accurately categorizing images of double-handed gestures into ten different classes.

Telugu Data Classification

Abstract: Telugu is considered as the difficult languages which is morphologically rich when it comes to Dravidian languages. There are many Telugu documents available on Internet, it is important to organize the data by automatically by assigning a collection of text with predefined categories. Here the Telugu data is classified into multiple areas like business, sports, entertainment, nation, editorial is the main goal throughout this research work. This research work provides up an efficient model by adopting some ML classifiers such as SVM, Naive Bayes, and Logistic regression to perform some areas of classification on Telugu data. The results obtained by various machine-learning models are compared and an efficient model is discovered, and it is observed that the Naive Bayes model outperformed with reference to accuracy, precision, recall, and F1- score

“Deep Learning Based Classification of Single-Hand South Indian Sign Language Gestures”

Gesture recognition is a branch of computer science and language technology dedicated to utilizing mathematical algorithms for the analysis of human gestures. Within the realm of non-verbal communication, the pivotal role of human arm movements and gestures remains a focal point. This research introduces advanced multi-stream deep transfer learning models tailored for identifying signs from South Indian languages, specifically Kannada, Tamil, and Telugu. The primary aim is to offer support to individuals encountering speech disorders or disabilities. The key deep transfer learning models utilized include Inception-V3, VGG-16, and ResNet-50, which have been modified and improved to attain heightened classification efficacy. The dataset comprises 35,000 images capturing single-hand gestures. In the realm of models, Inception-V3 exhibits the utmost test accuracy, standing at 91.45%, alongside a validation accuracy of 93.45% when tasked with the classification of single-hand gesture images across thirty-five distinct categories. The importance of this study lies in its prospective utility for creating an automated system that can support and improve the functional capabilities of individuals facing speech disorders or disabilities.

A Glossary of Distress Expressions Among Kannada-Speaking Urban Hindu Women.

People's lived experiences of distress are complex, personal, and vary widely across cultures. So, too, do the terms and expressions people use to describe distress. This variation presents an engaging challenge for those doing intercultural work in transcultural psychiatry, global mental health, and psychological anthropology. This article details the findings of a study of common distress terminology among 63 Kannada-speaking Hindu women living in Mysuru, the second largest city in the state of Karnataka, South India. Very little existing scholarship focuses on cultural adaptation for speakers of Dravidian languages like Kannada; this study aims to fill this gap and support greater representation of this linguistic family in research on mental health, idioms of distress, and distress terminology. Between 2018 and 2019, we conducted a 3-phase study consisting of interviews, data reduction, and focus group discussions. The goal was to produce a non-exhaustive list of common Kannada distress terms that could be used in future research and practice to translate and culturally adapt mental health symptom scales or other global mental health tools.

The maternal U1 haplogroup in the Koraga tribe as a correlate of their North Dravidian linguistic affinity.

Introduction: The Koraga tribe are an isolated endogamous tribal group found in the southwest coastal region of India. The Koraga language shares inherited grammatical features with North Dravidian languages. To seek a possible genetic basis for this exceptionality and understand the maternal lineage pattern, we have aimed to reconstruct the inter-population and intra-population relationships of the Koraga tribal population by using mtDNA markers for the hypervariable regions along with a partial coding region sequence analysis. Methods and Results: Amongst the 96 individuals studied, we observe 11 haplogroups, of which a few are shared and others are unique to the clans Soppu, Oṇṭi and Kuṇṭu. In addition to several deep rooted Indian-specific lineages of macrohaplogroups M and U, we observe a high frequency of the U1 lineage (∼38%), unique to the Koraga. A Bayesian analysis of the U1 clade shows that the Koraga tribe share their maternal lineage with ancestral populations of the Caucasus at the cusp of the Last Glacial Maximum. Discussion: Our study suggests that the U1 lineage found in the Indian subcontinent represents a remnant of a post-glacial dispersal. The presence of West Asian U1 when viewed along with historical linguistics leads us to hypothesise that Koraga represents a mother tongue retained by a vanquished population group that fled southward at the demise of the Indus civilisation as opposed to a father tongue, associated with a particular paternal lineage.